Abstract
ABSTRACT
Gastric cancer and diabetes are two major global health challenges, and recent studies have indicated a potential association between these diseases. This study utilized NHANES database (2007–2016) to identify diseases linked to gastric cancer, with diabetes selected for further investigation. Mendelian Randomization analysis was applied to establish a causal relationship between the two conditions. To explore underlying mechanisms, mRNA data from the TCGA and GEO databases were analyzed using differential gene expression analysis, WGCNA, and machine learning techniques. Diagnostic models for gastric cancer and diabetes were developed. Analysis of NHANES data revealed a significant correlation between diabetes and gastric cancer. MR analysis suggested that diabetes may reduce the risk of gastric cancer. Differential gene expression analysis identified key genes associated with both diseases, while GO and KEGG enrichment analyses uncovered relevant biological pathways. WGCNA highlighted shared gene modules and core genes. Machine learning algorithms identified six diagnostic genes (DGs) and two prognostic genes for gastric cancer, with robust diagnostic models constructed and validated. This study demonstrates a potential link between gastric cancer and diabetes, offering new perspectives for disease management. The identified DGs and prognostic genes hold promise as tools for improving the diagnosis and prognosis of gastric cancer. This finding should be interpreted as hypothesis‐generating and requires experimental validation.